A canine model will be utilized, in an attempt to reproduce a progressive and fatal peritonitis which mimics the pathophysiology of severe clinical sepsis. As sepsis develops, frequent measurements will be made of cardiovascular, respiratory and oxygen utilization to (1) document the validity of the model and (2) relate these changes to effects on metabolism, catecholamines, coagulation, serum complement, liver and renal function. The dog model will be prepared by implanting a Swan-Ganz pulmonary arterial catheter and a systemic arterial catheter. After 48 hours, frozen, processed human feces mixed with barium sulfate will be injected into a gauze sponge in the peritoneal cavity. This produces peritonitis with death of the dog within 2-5 days. Morphine sulfate (5 mg i.m. q 4 hrs) is given for analgesia. Baseline and daily studies will be obtained for catecholamines, SMA-12, total serum complement, CBC, platelets, fibrinogen, serum osomolality, arterial-venous oxygen tension and content, carbon dioxide and pH. Urine creatine, osmolarity and hourly urine volume will be measured. Cardiac output, central venous pressure, systemic arterial pressure, pulmonary artery and pulmonary catheter wedge pressure will be obtained. Wedge pressures will be maintained through the study at 5 mmHg. Control animals will be sacrificed on the fifth study day. All animals will have post-mortem examinations. An attempt will be made to modify the pathophysiology with prostaglandin cyclo-oxygenase blocker, vasoactive agents and pharmacologic doses of methylprednisone. Localization of platelet sequestration will be undertaken with 111-Indium labeled platelets.